Romania Cut Electricity Imports to 1.8% as Renewables Supplied 52% of Production. The Next Bottleneck Is Export Capacity

The headline number is real. What it measures is not energy independence. It is a country that has learned to sell cheap and buy expensive on the same day, and the constraint on fixing that is now the grid, not the generation fleet.

Read the word ‘net’ before you read anything else

In the first five months of 2026, Romania’s electricity production rose 6.2 percent year on year to 22.3 TWh while internal consumption fell 1.6 percent to 22.7 TWh. Net imports collapsed fivefold to 0.4 TWh, or 1.8 percent of internal consumption, down from 9.0 percent in the same period of 2025. Production from hydro, solar and wind grew 23 percent and covered 52 percent of total output, with solar up 35 percent.

Every one of those figures is accurate. Together they are being read as a story about self-sufficiency, and that reading does not survive the next line of the same dataset.

Gross imports fell only 11.8 percent, to 6.9 TWh. They still equalled 30.3 percent of total internal consumption. Exports rose 13.1 percent to 6.5 TWh. Romania did not stop importing electricity. It started exporting almost as much as it imports, and the two numbers happened to cancel out.

That is not independence. That is churn, and churn has a price.

The value asymmetry is the whole trade

The two flows do not happen at the same hours or at the same prices. Romania imports at expensive peak hours and exports during the middle of the day, when solar output is at its highest and prices are at their lowest. That combination puts upward pressure on domestic prices even when the annual TWh balance looks healthy.

So a country can be roughly balanced in energy and structurally negative in money. For anyone trading the Romanian border, that gap is not a footnote. It is the position.

One more nuance that gets lost in the headline. A meaningful share of Romanian exports goes to Moldova, which is not surplus solar being dumped at midday but a genuine supply relationship. Moldova consumes roughly a tenth of what Romania does and already covers up to half of that domestically. Strip the Moldovan volume out and the remaining export book is weighted even more heavily toward cheap daytime solar.

What actually produced the 52 percent

Renewable growth is genuine. Hydro, solar and wind all posted strong gains and solar in particular is compounding fast. But a share is a ratio, and three things moved the denominator in the same direction.

• Nuclear was down. A planned outage at Cernavoda held output to 4 TWh, 12.3 percent lower year on year. Nuclear is roughly a fifth of Romanian production in a normal year, so a double-digit fall mechanically lifts every other share.
• Thermal was used less, down 4.4 percent, partly a merit-order effect and partly a fuel-cost effect.
• Consumption fell. End-user consumption dropped 3.4 percent to 20.2 TWh, with the residential segment down 11.3 percent to 4.5 TWh and the economy down 0.8 percent to 15.5 TWh. Some of the residential decline is a measurement artefact, since prosumer self-consumption is not counted in production statistics, but a shrinking industrial and commercial load is not something to celebrate.

 

A 52 percent renewable share achieved partly through weaker nuclear availability and weaker demand is a different asset than a 52 percent share achieved through growth alone. Both are worth having. Only one of them is durable.

August 2026 tested the thesis, and the thesis moved

The 1.8 percent figure was published on 20 July, covering January to May. Within four weeks it looked like a description of a different country.

Record low Danube flows removed the cooling water Cernavoda needs. Unit 1 came offline on 28 July, Unit 2 followed in mid-August, and Romania’s only nuclear plant, normally around 20 percent of national production, went to zero. The Danube’s inflow was forecast at roughly 1,350 to 1,400 cubic metres per second against a multi-year August average near 3,900. Romania declared an energy sector state of alert, issued an early warning on 7 August and notified the European Commission on 10 August under Article 14(2) of the EU electricity risk-preparedness regulation. Coal retirements were postponed, including a Turceni unit due to close on 31 August under the national recovery plan. Imports from Bulgaria, Hungary, Serbia and Ukraine rose. Dacia and Ford suspended production to take industrial load off the system.

The Ministry of Energy was clear that supply was manageable and equally clear that prices would rise to reflect costly imports. That is the honest version of Romanian energy security in 2026: physically resilient, financially exposed, and resilient largely because the interconnectors held.

Which is the point. The same infrastructure that let Romania survive August is the infrastructure that determines whether it can monetise a surplus in May.

Why export capacity is the next bottleneck, in four layers

Export capacity is not one problem. It is four, and only the first is made of steel.

• Physical evacuation. Transelectrica’s ten-year transmission development plan is worth more than 2 billion euros. The 2026 investment programme is 914 million lei, more than 30 percent above the 691 million lei spent in 2025. The projects that matter for exports are specific: 400 kV upgrades at Stâlpu and Teleajen to move Dobrogea output toward Muntenia, the Timișoara to Arad 400 kV line, the Porțile de Fier node, and the Suceava to Bălți line that integrates Moldova into ENTSO-E. Transelectrica expects roughly 5,000 MW of additional integration capacity in Dobrogea and Banat by 2027.
• Commercial capacity. Physical wires are not the same as capacity offered to the market. EU rules require transmission operators to make at least 70 percent of interconnector capacity available for cross-zonal trade, the MACZT rule, with a deadline that has already passed. ACER’s monitoring has repeatedly found the target far from met across the bloc. A gigawatt of copper that is not offered to the day-ahead market does not export anything.
• Methodology. Central European Core markets operate under flow-based market coupling. Southeast Europe still runs on net transfer capacity. A March 2026 study published in Academia Green Energy on causal price spillovers found that the Core forms a more integrated system with faster price convergence, while Southeast European markets show weaker causal connectivity and slower adjustment. In practical terms, Romania and its neighbours are operating a partly decoupled regulatory regime, and that decoupling is worth real money in the wrong direction.
• Legislation. On 26 June 2026 EU energy ministers agreed the Council position on the European grids package, covering a TEN-E revision and a permitting acceleration directive. The most commercially interesting provision is the recycling of congestion income into cross-border projects, starting at 10 percent from 1 January 2028 and rising to 25 percent by 2031. The package also treats electricity and renewable projects as being of overriding public interest and allows member states to deem silence from authorities as tacit approval. Trilogue with Parliament is targeted to conclude in 2026.

 

The cost of getting these wrong is measurable. On 10 July 2024, day-ahead prices across the region tracked each other through the morning and then decoupled sharply between 19:00 and midnight: Hungary reached roughly 693 euros per MWh, Romania around 442 euros, while Austria stayed below 200. The study attributes that divergence to reduced transfer availability on interfaces including Romania to Bulgaria and Greece to Bulgaria. Two hundred and fifty euros per MWh of spread between neighbours is not a generation problem.

The pipeline will make this harder before it makes it easier

Romania passed 2,500 MW of new generation and storage connected to the grid in 2026 alone, reaching 2,514 MW with the connection of the IS Părău storage facility at 76.76 MW and 150.288 MWh. Transelectrica expects roughly another 1,000 MW of solar and wind plus around 1,000 MW of storage by year end, alongside new gas capacity at Mintia and Iernut.

Offshore is queued behind it. In July 2026 the Ministry of Energy published six Black Sea zones totalling 11.5 GW for public consultation, phased, with a first tranche of 3.1 GW covering one 1,900 MW perimeter and two of 600 MW each, aimed at the 3 GW by 2035 target in the Romanian Energy Strategy 2025-2035. The World Bank’s roadmap puts Romania’s technical offshore potential at roughly 76 GW, around 22 GW of it fixed-bottom, with average wind speeds above 7.5 metres per second.

Every megawatt of that lands on the same midday hours that already push Romanian prices toward zero. Without evacuation and cross-zonal capacity, additional generation does not become additional revenue. It becomes additional curtailment and additional negative-price exposure.

Why Romania is the right place to build for export

The location argument is stronger than the current numbers suggest, and it rests on geography plus institutions rather than optimism.

• Five borders. Romania is synchronously interconnected with Hungary, Bulgaria, Serbia and Ukraine, and supplies Moldova. Very few EU markets sit on that many interfaces, and each one is a separate price surface to trade against.
• A growing non-EU offtake market next door. The Suceava to Bălți 400 kV line is being built specifically to integrate Moldova into ENTSO-E. Moldova is small, but it is a structurally short market directly adjacent to Romanian generation, and it is being wired into the European system through Romanian infrastructure.
• A new corridor to the Caucasus with Romania as the landing point. The Black Sea submarine cable from Constanța to Anaklia received Project of Mutual Interest status from the European Parliament and Council in April 2026, and Transelectrica signed a memorandum of understanding with Georgian State Electrosystem, moving the project into technical preparation in July. The link is planned at 1,155 km, of which about 1,115 km is submarine, with capacity up to 1,300 MW. If it is built, Romania is not a peripheral EU market. It is a gateway.
• Resource depth that no other EU Black Sea state has. Romania is the only EU member with meaningful Black Sea offshore wind potential at this scale, and offshore wind produces at hours that onshore solar does not, which improves the export profile rather than worsening it.
• Evacuation is being funded, not just discussed. A 30 percent increase in the transmission capital programme and a targeted 5,000 MW of additional integration capacity in Dobrogea and Banat by 2027 is a specific, dated commitment aimed at exactly the two regions where the renewable build-out is concentrated.
• The EU money will follow the congestion. The congestion income recycling mechanism agreed in the Council position channels funds toward reducing cross-border bottlenecks from 2028. Romania sits on congested interfaces. A country with expensive congestion is, under this design, a country with a funding claim.
• Scarcity cuts both ways. The evening scarcity that makes Romania an importer is the same scarcity that exists across Southeast Europe at the same hour. Any Romanian asset that can deliver firm or stored energy into that window is selling into a regional shortage, not just a domestic one.

 

Romania will not become a net exporter because it generates more. It will become one, if it does, because it can move and time what it already generates.

The case against, stated properly

• The 1.8 percent is a five-month figure measured against falling consumption, and August 2026 has already tested it to destruction.
• Demand contraction is not a foundation. Residential consumption down 11.3 percent and industrial demand flat to negative means the export position partly reflects a weaker economy.
• Hydrology risk is now systemic rather than seasonal. Low Danube flows simultaneously hit nuclear cooling, hydro output, thermal plant operation and fuel logistics. That is correlated risk across the entire dispatchable fleet.
• Offshore wind has a credibility problem. Law 121/2024 deadlines for perimeters, concession rules and support schemes were missed, and there is still no published auction timetable. Zoning for consultation is a real step, but first steel in water remains a 2030s event.
• The Black Sea cable is at memorandum stage. PMI status and technical preparation are meaningful, but a 1,155 km submarine link is a decade-scale project with no final investment decision and no announced financing package.
• MACZT compliance is not in Romania’s sole gift. Cross-zonal capacity is set jointly with neighbouring operators, and unilateral capacity reductions on the other side of a border constrain Romanian exports regardless of what Transelectrica does.

 

What to watch

1. Whether the second half of 2026 reverses the net import trend. Given the nuclear outage, a return toward mid-single-digit net import dependence would be the base case, and the full-year figure will settle the argument.
2. Romania to Hungary and Romania to Bulgaria MACZT values in the next ACER monitoring cycle. That single dataset says more about export potential than any capacity announcement.
3. The timetable for Southeast European migration from net transfer capacity to flow-based capacity calculation. This is the highest-leverage unglamorous reform on the list.
4. The grids package trilogue outcome, and whether the congestion income percentages survive negotiation with Parliament intact.
5. Transelectrica’s grid capacity auction results and the regional split of available capacity, which reveals where evacuation actually exists rather than where projects want to be.
6. The offshore wind concession tender timetable and the shape of the support scheme, which determines whether 3.1 GW is a plan or a press release.
7. Black Sea cable milestones: financing structure, CEF applications and any move from memorandum toward investment decision.
8. Danube hydrology and whether coal retirement dates slip again. Every postponement is a signal about how much firm capacity the system actually believes it has.

 

Momentum Energy’s View

The 1.8 percent is a real milestone and a misleading headline, and both things need saying in the same sentence. Romania has not achieved energy independence. It has achieved a temporary volumetric balance while remaining structurally exposed on price, and August 2026 demonstrated how quickly that balance inverts when one large unit goes offline.

The metric that matters is not the TWh balance. It is the euro spread between the hours Romania exports and the hours Romania imports. On that metric Romania is not close to balanced, and no amount of additional solar fixes it. More midday generation into a constrained network makes the spread worse, not better.

That reframes the investment thesis. The opportunity in Romania is not generation capacity. It is time-shifting and space-shifting: storage that moves energy across the sunset, firm capacity that survives a drought, and cross-border capacity that moves surplus to a market willing to pay for it. Three of those four bottlenecks are regulatory rather than physical, which is unusual, and unusually tractable.

On location, our reading of the evidence favours Romania. Five interconnected neighbours, a Moldovan market being wired into ENTSO-E through Romanian lines, a Black Sea corridor to the Caucasus carrying EU Project of Mutual Interest status, roughly 76 GW of technical offshore potential, a transmission capital programme up 30 percent, and an EU mechanism from 2028 that funnels congestion revenue toward exactly the bottlenecks Romania has. Few markets in the region hold that combination.

What Romania lacks is not resource, partners or capital appetite. It lacks offered cross-zonal capacity, a harmonised capacity calculation methodology with its neighbours, and firm evacuation into Dobrogea and Banat ahead of the connection queue rather than behind it. Those are decisions, not constraints.

For traders, the near-term edge is in the shape, not the balance, and specifically in the interfaces where transfer availability is restricted during evening stress. For independent power producers, the near-term risk is assuming a grid connection agreement equals evacuation certainty. For infrastructure investors, the durable position is anything that shortens the distance between a cheap Romanian midday and an expensive regional evening.

Romania spent a decade being described as an import-exposed market. It is now a market with a surplus it cannot always sell. That is a better problem, and a solvable one, but it is not the same as being solved.